The use of genome-wide expression studies is revolutionizing our approach to understanding normal cellular processes and the mechanisms of disease. Sophisticated computational methods enable the analysis of this data. An important goal is to empower scientists with domain expertise by giving them direct access to these new technologies and techniques for interpretation. However, these computational methods can be difficult to understand and use correctly. They may not easily work together or be reproduced. Over the past 6 years we have been developing advanced mathematical methods and computational algorithms for the analysis of microarray data. We share them with the research community in user-friendly, freely available software packages, GeneCluster and GenePattern. GenePattern was released in March 2004 and represents a major reworking, improvement, and expansion of the GeneCluster application. The power of GenePattern is its accessibility to a broad community of users, the ability to access and interoperate a library of analytic and visualization modules, and the ease with which the environment supports the rapid development and dissemination of new methods. Our goal is to continue to make this software available to and useful for the research community. Aim 1. Continuing user support for the GenePattern and GeneCluster packages. Aim 2. Enhancements to GenePattern to better support users and their research. Aim 3. Continuing software maintenance of the GenePattern and GeneCluster Packages. Our extensive experience in software engineering, significant user base, development of preliminary documentation, and successful workshops for users make us well poised to carry out the aims of this proposal.